Data-driven modeling of the neural dynamics underlying language processing

In this thesis, we use data-driven techniques in an attempt to explain the neural data during speech and language perception in unconstrained naturalistic setup. We aim to use a combination of data exploration approaches and data-driven feature extraction models to provide a more bottom-up way of studying the neural responses. We hope that these data-driven strategies can confirm some of the previous theory-based neurolinguistic findings but at the same time offer new insight regarding how the human brain processes speech related information.